What We'll Cover
We will take you through the ins and outs of our sports betting analytics system and predictive techniques behind our algorithm.
We pull back the curtain on deceptive marketing tactics and show why units returned, not win percentage, is the deciding factor in making a profit.
While we use a variety of sports betting analytics tools, we will explore: individual player-level ability, team-level performance, and team-level play style.
We tie everything together to show you how our sports betting strategy uses proprietary metrics to create a revolutionary algorithm.
We designed our system to maximize units returned in the betting industry using predictive analytics and sports data.
Units returned is the single most important factor in becoming a successful sports bettor. Units returned translates into profit, which generates a return on your investment.
There’s a lot of false marketing in the industry that showcases win percentage as the most important factor in sports betting. While win percentage is important, we must consider the method behind accurate predictions and understand that not all wins are the same.
This is where we break the stigma. We are fully transparent in everything we do and answer the most important question in sports betting:
“Am I making money?”
SBIA's units returned from the last 5 years in the NBA.
" -" represents covid effected months.
" / " represents non wagered months.
Let's Apply This Theory
Let's take 10 games using the moneyline as an example. If you were to have a 7-3 record, that is a 70% win percentage. This is considered a phenomenal win percentage, but the odds wagered for these 10 games matter. Let’s say that you have a $400 unit size, and all 10 games you wagered were at -200 odds. With a 7-3 record you would have generated .5 units, which is a profit of $200.
Total Amount Wagered:
7 Games Won:
3 Games Lost:
$4,000 ($400 x 10)
$1,400 ($400 x 7 / 2)
- $1,200 ($400 x -3)
$200 ($1,400 - $1,200)
.5 unit ($200 / $400)
SBIA's percentage of games wagered within a single NBA season.
Percentage of games wagered
Play the Numbers
Our methodology maximizes unit return by utilizing a large volume of plays. A large volume of plays has two primary benefits: stable performance and bankroll turnover.
We treat sports betting similar to how the S&P 500 approaches investments. The larger the sample size, the more stable the market. By wagering on a larger volume of games, we decrease the volatility that is involved in this environment.
This allows us to build a diverse portfolio for our clients using data analytics, stabilizing long-term performance through a difficult stretch in the market.
Turn Up the Volume
The second benefit for a large volume of plays is bankroll turnover. While we provide a diversification effect similar to the S&P 500, our advantage here is that a traditional investment will take your $10,000 and provide an annual 17% return, leaving you with $11,700. At SBIA, we use data points, artificial intelligence, and machine learning methods to improve our strategy.
Volume and short-term expirations provide the opportunity to turn over your bankroll effectively, investing far more than the initial capital required to enter the market. At a 2% unit and 4,000 picks a year, you have turned over your bankroll 80 times, investing $800,000. Equivalently, we only have to produce a 0.21% return per dollar invested to generate the same $1,700 profit the S&P 500 did.
The reality is that we can generate 3-5% per dollar, say $32,000, which is a 320% ROI. Volume gives us an opportunity to turn a small initial capital investment into a large yielding investment. This is something that traditional markets cannot provide.
Now that you understand why our methodology utilizes volume, it is vital to understand the importance of bankroll management.
Bankroll management is the key to sustainable sports betting. Our system can deliver a large volume of picks on any given gameday.
In order to place all of the given wagers, it is important to effectively manage your bankroll. A unit size of 2.5% provides the stability to mitigate any short term volatility.
Player -Level Ability
For individual player-level ability, we track a player’s improvement or decline in performance over time by monitoring the outcome of every play in which a player is in a game and updating our valuation of their abilities from each possession.
Let’s take a look at LeBron James's player valuations over two seasons:
In 2020-2021, James played only 45 games due to injury, averaging 25.0 points, 33.4 minutes, & had a 51.3% field goal percentage. While James was limited in his time on the court, our metrics showed he had a relatively high offensive and defensive valuation throughout the season.
Compared to the 2021-2022 season, James played 56 games, averaging 30.3 points, 37.2 minutes, and had a 52.4% field goal percentage. Despite playing more games and averaging better statistics, our betting data showed that he was less effective than the previous year. This is a testament to how we utilize individual player-level ability to uniquely analyze data.
LeBron James (2020 -2021)
LeBron James (2021 -2022)
Team -Level Performance
Monitoring team-level performances is very similar. Game results are the most basic indicator of a team’s ability. However, to pick up on the long-term changes that are not the product of chance by two teams, we consider the randomness of these results.
We can see this in the 2021-2022 NBA ratings for the Grizzlies, Nets, and Celtics.
You can see the result of the Nets consistent off-the-court and injury issues for most of the season, the Grizzlies breakout season behind their First Team All-Star, Ja Morant, and the Celtics going on a historic run in the second half of the season, despite their early struggles.
Team - Level Play Style
Pace is a commonly used metric to determine how fast a team has played in their previous games, which is approximately how many possessions each team gets in 48 minutes.
The major flaw with pace, however, is it tells you how fast a game was played, but not how fast each individual team attempts to play the game. This traditional pace metric can fail when you use it to predict a future matchup.
For this reason we create a team specific pace metric that dials in on each team’s speed of play. This can then be used to better predict the pace of future matchups with different opponents.
Now that we have explored a few of our metrics, we plug this data into our betting system to project win probabilities, point spreads, and point totals for each team. We compare these projections to the available odds and lines against the spread, totals, and moneyline markets to determine an expected value of each option.
Expected value is the measure of how much we expect to win if a game was repeated an infinite number of times. If the expected value is positive, that’s a game we will play.
However, we live in a random world where each event is only observed once and the outcome is binary. An individual game cannot provide the expected value alone. To achieve this, we rely on volume, allowing our data science to prevail.
Money by the Markets
Our model identifies positive expected value bets across several markets: first quarter, first half and full game moneyline, first half and full game spreads, totals and team totals. We could provide you picks in all seven, none, or any combination of these markets on any given gameday.
First Half Moneyline
SBIA's overall sides market results from the 2017 -2021 NBA seasons.
First Quarter Moneyline
First Half Spread
Full Game Moneyline
Full Game Spread
All Side Markets
Units Units Units
Depending on the expected value of a game, we allocate 1 unit, 1.5 units, or 2.0 units to each pick. The number of picks we have over correlated markets is also considered in how many units will be allocated.
For example, if we have picks on the full game moneyline and the full game against the spread, we will have a max of 1 unit each since the outcomes of those markets are highly correlated.
SBIA's pick delivery example.